Hello,
Recently needed to install different RAM due to an issue which seems to have creeped up with my old one.
Topic for that here.
New RAM: Samsung 16GB ECC DualRank modules x4
My RAM went in and the memory training was happening and I set the few things in the BIOS to the RAM settings and set everything else to Auto. (Just set 1.22V due to possible voltage drop and set the primary timings to JEDEC default for 3200)
So now every time I start the PC cold it seems to have to re-train the memory so PC shuts down once and starts up again and everything works.
I don’t know though if this is normal for ECC RAM or if my board is just a dumb board?
I really thought RAM compatibility got better with Zen 2 it feels not like it honestly.
So memory trains different when it’s hot vs cold. It can. So this is not super a typical. I don’t see anything unsettling in your timings so far. You could use aida64 and memory benchmark to confirm no issue. It’ll be easy to spot if the bandwidth is way different
Interesting so that behaviour can happen, I never had it happen on all the systems I’d had so far so I was very suprised to see it.
So it’s not something specific to ECC if I interpreted it right?
AIDA64 tells me the following:
I don’t know if good or bad, but it seems normal enough for me.
Just as a joke I tried to see if the memory would boot with 3600 and to my surprise it did it, I’m not running it but was fun to see. Don’t want to push my luck here as I prefer a stable system over the little bit I get with OCing it.
.